At present, little is known about the process of clonal evolution in premalignant lesions or how it is associated with the risk of progression to cancer. There is also a shortage of techniques that can be used to effectively monitor high-risk patients to prevent cancer development or, in patients with a cancer history, the local recurrence of a cancer. The objective of the proposed study is to obtain information on the temporal pattern of clonal change in high-risk oral lesions using exfoliative cells. The hypothesis to be tested is that the evaluation of allelic loss in exfoliative cells collected from scrapes of lesions will be an independent predictor of risk of progression of dysplastic lesions in non-cancer patients. A second hypothesis is that data obtained from scrapes of former cancer sites will predict clinical outcome of oral cancer patients during follow-up. In all cases, the association of allelic loss and outcome will be assessed independently for scrapes and biopsies (gold standard). The study will involve 50 patients with oral dysplasia and 100 patients with SCC. Samples will be obtained from the primary lesion (dysplasia or SCC) of each patient and then collected at 6-month intervals for 2 years. All samples will be assayed for allelic loss at 8 chromosome arms, chosen on the basis of our prior results with archival samples: 3p, 4q, 8p 9p, 11q, 13q, 14q, and 17p. The significance of the study is that if the pattern of allelic loss in scrapes is predictive of clinical outcome, this noninvasive approach could be used by clinicians to indicate which patients would require more aggressive treatment as well as to monitor the success of such treatments (i.e. disappearance of genetic clones). Possible treatments could include more traditional forms of therapy (e.g. surgery) or the use of innovative approaches with chemopreventive agents focused specifically on the genetic clones.